In [71]:
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
import statsmodels.api as sm
from sklearn.preprocessing import LabelEncoder
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
In [2]:
df = pd.read_csv(r"C:\Users\User\OneDrive\Documents\4. Github Project & Dataset\python project\sorted_data.csv")
df.head()
Out[2]:
loan_paid_back education_level monthly_income employment_status debt_to_income_ratio credit_score loan_amount interest_rate loan_term installment num_of_open_accounts total_credit_limit current_balance delinquency_history public_records num_of_delinquencies
0 1 Master's 2020.02 Employed 0.07 743 17173.72 13.39 36 581.88 7 40833.47 24302.07 1 0 1
1 1 Bachelor's 1681.08 Employed 0.22 531 22663.89 17.81 60 573.17 5 27968.01 10803.01 1 0 3
2 1 High School 2181.82 Employed 0.23 779 3631.36 9.53 60 76.32 2 15502.25 4505.44 0 0 0
3 1 High School 989.49 Employed 0.26 809 14939.23 7.99 36 468.07 7 18157.79 5525.63 4 0 5
4 1 Other 2110.54 Employed 0.26 663 16551.71 15.20 60 395.50 1 17467.56 3593.91 2 0 2
In [3]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 20000 entries, 0 to 19999
Data columns (total 16 columns):
 #   Column                Non-Null Count  Dtype  
---  ------                --------------  -----  
 0   loan_paid_back        20000 non-null  int64  
 1   education_level       20000 non-null  object 
 2   monthly_income        20000 non-null  float64
 3   employment_status     20000 non-null  object 
 4   debt_to_income_ratio  20000 non-null  float64
 5   credit_score          20000 non-null  int64  
 6   loan_amount           20000 non-null  float64
 7   interest_rate         20000 non-null  float64
 8   loan_term             20000 non-null  int64  
 9   installment           20000 non-null  float64
 10  num_of_open_accounts  20000 non-null  int64  
 11  total_credit_limit    20000 non-null  float64
 12  current_balance       20000 non-null  float64
 13  delinquency_history   20000 non-null  int64  
 14  public_records        20000 non-null  int64  
 15  num_of_delinquencies  20000 non-null  int64  
dtypes: float64(7), int64(7), object(2)
memory usage: 2.4+ MB
In [4]:
df.value_counts("education_level")
Out[4]:
education_level
Bachelor's     8045
High School    5919
Master's       3724
Other          1508
PhD             804
Name: count, dtype: int64
In [5]:
df.value_counts("employment_status")
Out[5]:
employment_status
Employed         13007
Self-employed     2923
Unemployed        2113
Retired           1176
Student            781
Name: count, dtype: int64
In [6]:
df.describe()
Out[6]:
loan_paid_back monthly_income debt_to_income_ratio credit_score loan_amount interest_rate loan_term installment num_of_open_accounts total_credit_limit current_balance delinquency_history public_records num_of_delinquencies
count 20000.000000 20000.000000 20000.000000 20000.00000 20000.000000 20000.000000 20000.00000 20000.000000 20000.000000 20000.000000 20000.000000 20000.000000 20000.000000 20000.000000
mean 0.799900 3629.136466 0.177496 679.25695 15129.300909 12.400627 43.22280 455.625794 5.011800 48649.824769 24333.394631 1.990150 0.061800 2.489150
std 0.400085 2389.048326 0.105108 69.63858 8605.405513 2.442729 11.00838 274.622125 2.244529 32423.378128 22313.845395 1.474945 0.285105 1.631384
min 0.000000 500.000000 0.010000 373.00000 500.000000 3.140000 36.00000 9.430000 0.000000 6157.800000 496.350000 0.000000 0.000000 0.000000
25% 1.000000 2021.730000 0.100000 632.00000 8852.695000 10.740000 36.00000 253.910000 3.000000 27180.492500 9592.572500 1.000000 0.000000 1.000000
50% 1.000000 3048.770000 0.160000 680.00000 14946.170000 12.400000 36.00000 435.595000 5.000000 40241.615000 18334.555000 2.000000 0.000000 2.000000
75% 1.000000 4556.495000 0.240000 727.00000 20998.867500 14.002500 60.00000 633.595000 6.000000 60361.257500 31743.327500 3.000000 0.000000 3.000000
max 1.000000 33333.330000 0.670000 850.00000 49039.690000 22.510000 60.00000 1685.400000 15.000000 454394.190000 352177.900000 11.000000 2.000000 11.000000
In [7]:
print(f"Minimum loan Amount: {min(df["loan_amount"])}")

Minimum loan Amount: 500.0
In [8]:
print(f"Maximum loan Amount: {max(df["loan_amount"])}")

Maximum loan Amount: 49039.69
In [9]:
print(df.isna().sum())

loan_paid_back          0
education_level         0
monthly_income          0
employment_status       0
debt_to_income_ratio    0
credit_score            0
loan_amount             0
interest_rate           0
loan_term               0
installment             0
num_of_open_accounts    0
total_credit_limit      0
current_balance         0
delinquency_history     0
public_records          0
num_of_delinquencies    0
dtype: int64
In [10]:
df.drop_duplicates()
Out[10]:
loan_paid_back education_level monthly_income employment_status debt_to_income_ratio credit_score loan_amount interest_rate loan_term installment num_of_open_accounts total_credit_limit current_balance delinquency_history public_records num_of_delinquencies
0 1 Master's 2020.02 Employed 0.07 743 17173.72 13.39 36 581.88 7 40833.47 24302.07 1 0 1
1 1 Bachelor's 1681.08 Employed 0.22 531 22663.89 17.81 60 573.17 5 27968.01 10803.01 1 0 3
2 1 High School 2181.82 Employed 0.23 779 3631.36 9.53 60 76.32 2 15502.25 4505.44 0 0 0
3 1 High School 989.49 Employed 0.26 809 14939.23 7.99 36 468.07 7 18157.79 5525.63 4 0 5
4 1 Other 2110.54 Employed 0.26 663 16551.71 15.20 60 395.50 1 17467.56 3593.91 2 0 2
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
19995 0 Bachelor's 3303.34 Employed 0.28 691 16322.23 15.05 36 566.22 2 23748.10 5801.45 1 0 4
19996 1 Bachelor's 2671.91 Employed 0.37 758 16697.34 11.89 36 553.71 8 49929.65 40901.31 3 0 3
19997 1 Master's 1553.50 Student 0.11 751 23924.78 10.06 36 772.66 3 13137.57 5075.67 1 0 2
19998 1 Master's 1848.45 Retired 0.28 646 16920.13 16.06 36 595.36 5 19580.82 3876.16 4 0 5
19999 0 Other 1978.14 Employed 0.23 630 15769.75 13.07 36 531.88 8 43013.59 12753.03 2 0 2

20000 rows × 16 columns

In [11]:
plt.figure(figsize=(10,6))
sns.countplot(df, x="loan_paid_back",color = 'orange')
plt.title("Count Plot of Loan Paid Back")  
plt.xlabel("Loan Paid Back")
plt.show()
No description has been provided for this image
In [12]:
plt.figure(figsize=(10,6))
sns.boxplot(df, x="monthly_income")
plt.title("Boxplot of Monthly Income")  
plt.xlabel("Monthly Income")
plt.show()
No description has been provided for this image
In [13]:
seventyfive_percentile_mothlyincome=df["monthly_income"].quantile(0.75)
In [14]:
tewntyfive_percentile_mothlyincome=df["monthly_income"].quantile(0.25)
In [15]:
IQR=seventyfive_percentile_mothlyincome-tewntyfive_percentile_mothlyincome
print(IQR)

2534.765
In [16]:
upper_outlier = seventyfive_percentile_mothlyincome + (1.5*IQR)
lower_outlier = tewntyfive_percentile_mothlyincome - (1.5*IQR)
print(upper_outlier, lower_outlier)

8358.6425 -1780.4175
In [17]:
no_outliers= df[(df["monthly_income"]>=lower_outlier) & (df["monthly_income"]<=upper_outlier)].copy()
print(no_outliers)

       loan_paid_back education_level  monthly_income employment_status  \
0                   1        Master's         2020.02          Employed   
1                   1      Bachelor's         1681.08          Employed   
2                   1     High School         2181.82          Employed   
3                   1     High School          989.49          Employed   
4                   1           Other         2110.54          Employed   
...               ...             ...             ...               ...   
19995               0      Bachelor's         3303.34          Employed   
19996               1      Bachelor's         2671.91          Employed   
19997               1        Master's         1553.50           Student   
19998               1        Master's         1848.45           Retired   
19999               0           Other         1978.14          Employed   

       debt_to_income_ratio  credit_score  loan_amount  interest_rate  \
0                      0.07           743     17173.72          13.39   
1                      0.22           531     22663.89          17.81   
2                      0.23           779      3631.36           9.53   
3                      0.26           809     14939.23           7.99   
4                      0.26           663     16551.71          15.20   
...                     ...           ...          ...            ...   
19995                  0.28           691     16322.23          15.05   
19996                  0.37           758     16697.34          11.89   
19997                  0.11           751     23924.78          10.06   
19998                  0.28           646     16920.13          16.06   
19999                  0.23           630     15769.75          13.07   

       loan_term  installment  num_of_open_accounts  total_credit_limit  \
0             36       581.88                     7            40833.47   
1             60       573.17                     5            27968.01   
2             60        76.32                     2            15502.25   
3             36       468.07                     7            18157.79   
4             60       395.50                     1            17467.56   
...          ...          ...                   ...                 ...   
19995         36       566.22                     2            23748.10   
19996         36       553.71                     8            49929.65   
19997         36       772.66                     3            13137.57   
19998         36       595.36                     5            19580.82   
19999         36       531.88                     8            43013.59   

       current_balance  delinquency_history  public_records  \
0             24302.07                    1               0   
1             10803.01                    1               0   
2              4505.44                    0               0   
3              5525.63                    4               0   
4              3593.91                    2               0   
...                ...                  ...             ...   
19995          5801.45                    1               0   
19996         40901.31                    3               0   
19997          5075.67                    1               0   
19998          3876.16                    4               0   
19999         12753.03                    2               0   

       num_of_delinquencies  
0                         1  
1                         3  
2                         0  
3                         5  
4                         2  
...                     ...  
19995                     4  
19996                     3  
19997                     2  
19998                     5  
19999                     2  

[19076 rows x 16 columns]
In [18]:
print(no_outliers["monthly_income"].describe())

count    19076.000000
mean      3275.431041
std       1679.810750
min        500.000000
25%       1981.797500
50%       2940.785000
75%       4288.657500
max       8357.400000
Name: monthly_income, dtype: float64
In [19]:
plt.figure(figsize=(10,6))
sns.histplot(df, x="monthly_income",bins=50 ,color = 'green', kde=True)
plt.title("Histogram of Monthly Income with outliers")
plt.xlabel("Monthly Income")
plt.ylabel("Count")
Out[19]:
Text(0, 0.5, 'Count')
No description has been provided for this image
In [20]:
plt.figure(figsize=(10,6))
sns.histplot(no_outliers, x="monthly_income",bins=50 ,color = 'green', kde=True)
plt.title("Histogram of Monthly Income without outliers")
plt.xlabel("Monthly Income")
plt.ylabel("Count")
Out[20]:
Text(0, 0.5, 'Count')
No description has been provided for this image
In [21]:
plt.figure(figsize=(10,6))
sns.histplot(no_outliers, x="loan_amount",bins=50 ,color = 'red', kde=True)
plt.title("Histogram of Loan amount")
plt.xlabel("Loan Amount")
plt.ylabel("Count")
Out[21]:
Text(0, 0.5, 'Count')
No description has been provided for this image
In [22]:
df["loan_amount"].describe()
Out[22]:
count    20000.000000
mean     15129.300909
std       8605.405513
min        500.000000
25%       8852.695000
50%      14946.170000
75%      20998.867500
max      49039.690000
Name: loan_amount, dtype: float64
In [23]:
plt.figure(figsize=(10,6))
sns.histplot(df, x="credit_score",bins=50 ,color = 'blue', kde=True, line_kws={'color': 'black'})
plt.title("Histogram of Credit Score")
plt.xlabel("Credit Score")
plt.ylabel("Density")
Out[23]:
Text(0, 0.5, 'Density')
No description has been provided for this image
In [24]:
skewness = no_outliers.skew(numeric_only=True)
print(skewness)

loan_paid_back         -1.503481
monthly_income          0.780185
debt_to_income_ratio    0.789635
credit_score           -0.073550
loan_amount             0.255350
interest_rate           0.026442
loan_term               0.872613
installment             0.469447
num_of_open_accounts    0.442304
total_credit_limit      1.163923
current_balance         1.569846
delinquency_history     0.824823
public_records          5.018849
num_of_delinquencies    0.720254
dtype: float64
In [25]:
no_outliers["income_to_loan_ratio"]=((no_outliers["monthly_income"]*12)/(no_outliers["loan_amount"]))

no_outliers["utilization_ratio"]=((no_outliers["current_balance"]*12)/(no_outliers["total_credit_limit"]))
In [26]:
no_outliers['education_encoded'] =  LabelEncoder().fit_transform(no_outliers['education_level'])
In [27]:
no_outliers['employment_encoded'] =  LabelEncoder().fit_transform(no_outliers['employment_status'])
In [28]:
no_outliers.head()
Out[28]:
loan_paid_back education_level monthly_income employment_status debt_to_income_ratio credit_score loan_amount interest_rate loan_term installment num_of_open_accounts total_credit_limit current_balance delinquency_history public_records num_of_delinquencies income_to_loan_ratio utilization_ratio education_encoded employment_encoded
0 1 Master's 2020.02 Employed 0.07 743 17173.72 13.39 36 581.88 7 40833.47 24302.07 1 0 1 1.411473 7.141809 2 0
1 1 Bachelor's 1681.08 Employed 0.22 531 22663.89 17.81 60 573.17 5 27968.01 10803.01 1 0 3 0.890093 4.635157 0 0
2 1 High School 2181.82 Employed 0.23 779 3631.36 9.53 60 76.32 2 15502.25 4505.44 0 0 0 7.209927 3.487576 1 0
3 1 High School 989.49 Employed 0.26 809 14939.23 7.99 36 468.07 7 18157.79 5525.63 4 0 5 0.794812 3.651742 1 0
4 1 Other 2110.54 Employed 0.26 663 16551.71 15.20 60 395.50 1 17467.56 3593.91 2 0 2 1.530143 2.468972 3 0

no_outliers.to_csv("tranform_dataset.csv")

In [29]:
no_outliers.corr(numeric_only=True)
Out[29]:
loan_paid_back monthly_income debt_to_income_ratio credit_score loan_amount interest_rate loan_term installment num_of_open_accounts total_credit_limit current_balance delinquency_history public_records num_of_delinquencies income_to_loan_ratio utilization_ratio education_encoded employment_encoded
loan_paid_back 1.000000 0.019801 -0.222571 0.202053 -0.003618 -0.113965 -0.003623 -0.010736 0.003875 0.010068 0.003278 -0.083413 0.005399 -0.070747 0.005966 -0.005898 0.014446 -0.489946
monthly_income 0.019801 1.000000 -0.006746 0.005617 -0.003758 -0.005073 -0.006311 -0.003457 0.002780 0.842452 0.570038 0.003121 -0.001050 0.001463 0.212248 0.003826 -0.000182 -0.000054
debt_to_income_ratio -0.222571 -0.006746 1.000000 -0.021741 -0.004013 0.014475 0.005474 -0.002954 -0.005415 -0.001166 -0.001496 0.227326 -0.001707 0.209722 -0.005432 -0.002116 -0.005339 0.011981
credit_score 0.202053 0.005617 -0.021741 1.000000 0.007603 -0.569435 0.004303 -0.030968 -0.003292 -0.002704 -0.008101 -0.162587 0.003588 -0.142789 -0.001167 -0.011766 -0.001462 0.002165
loan_amount -0.003618 -0.003758 -0.004013 0.007603 1.000000 -0.006092 0.001611 0.945011 -0.005376 -0.001485 -0.008343 -0.006647 0.002117 -0.008046 -0.477663 -0.004393 -0.003510 0.010896
interest_rate -0.113965 -0.005073 0.014475 -0.569435 -0.006092 1.000000 -0.004115 0.060932 0.001876 0.000677 0.005138 0.091634 -0.003741 0.078712 0.010002 0.009797 0.001437 0.001008
loan_term -0.003623 -0.006311 0.005474 0.004303 0.001611 -0.004115 1.000000 -0.274990 0.015386 -0.002951 -0.000285 -0.002394 -0.003170 -0.002597 -0.002119 0.001168 0.008283 0.002489
installment -0.010736 -0.003457 -0.002954 -0.030968 0.945011 0.060932 -0.274990 1.000000 -0.009226 -0.000905 -0.007391 0.001220 0.004729 -0.001148 -0.450918 -0.004639 -0.007221 0.009685
num_of_open_accounts 0.003875 0.002780 -0.005415 -0.003292 -0.005376 0.001876 0.015386 -0.009226 1.000000 0.102172 0.066680 -0.000318 -0.008897 -0.003268 -0.008094 0.000687 0.010244 0.000554
total_credit_limit 0.010068 0.842452 -0.001166 -0.002704 -0.001485 0.000677 -0.002951 -0.000905 0.102172 1.000000 0.673074 0.002854 0.002314 0.003069 0.177502 -0.002482 -0.006075 -0.000430
current_balance 0.003278 0.570038 -0.001496 -0.008101 -0.008343 0.005138 -0.000285 -0.007391 0.066680 0.673074 1.000000 0.004341 -0.002354 0.008308 0.129031 0.648715 -0.002374 0.003713
delinquency_history -0.083413 0.003121 0.227326 -0.162587 -0.006647 0.091634 -0.002394 0.001220 -0.000318 0.002854 0.004341 1.000000 0.001683 0.903351 -0.002352 0.004545 0.003336 0.000028
public_records 0.005399 -0.001050 -0.001707 0.003588 0.002117 -0.003741 -0.003170 0.004729 -0.008897 0.002314 -0.002354 0.001683 1.000000 0.005856 -0.006165 -0.010095 0.008239 -0.008134
num_of_delinquencies -0.070747 0.001463 0.209722 -0.142789 -0.008046 0.078712 -0.002597 -0.001148 -0.003268 0.003069 0.008308 0.903351 0.005856 1.000000 -0.008850 0.005942 -0.000615 -0.004960
income_to_loan_ratio 0.005966 0.212248 -0.005432 -0.001167 -0.477663 0.010002 -0.002119 -0.450918 -0.008094 0.177502 0.129031 -0.002352 -0.006165 -0.008850 1.000000 0.003482 -0.005918 -0.005893
utilization_ratio -0.005898 0.003826 -0.002116 -0.011766 -0.004393 0.009797 0.001168 -0.004639 0.000687 -0.002482 0.648715 0.004545 -0.010095 0.005942 0.003482 1.000000 -0.000294 0.005680
education_encoded 0.014446 -0.000182 -0.005339 -0.001462 -0.003510 0.001437 0.008283 -0.007221 0.010244 -0.006075 -0.002374 0.003336 0.008239 -0.000615 -0.005918 -0.000294 1.000000 -0.009679
employment_encoded -0.489946 -0.000054 0.011981 0.002165 0.010896 0.001008 0.002489 0.009685 0.000554 -0.000430 0.003713 0.000028 -0.008134 -0.004960 -0.005893 0.005680 -0.009679 1.000000
In [30]:
plt.figure(figsize=(20,8))
sns.heatmap(no_outliers.corr(numeric_only=True), annot=True)
plt.title("Correlation between the numerical variables")
plt.show()
No description has been provided for this image
In [31]:
sns.pairplot(data=no_outliers)
plt.show()
No description has been provided for this image
In [32]:
X=no_outliers[["monthly_income", 
"debt_to_income_ratio",
"credit_score",
"loan_amount",
"interest_rate",
"loan_term",
"installment",
"num_of_open_accounts",
"total_credit_limit",
"current_balance",
"delinquency_history",
"public_records",
"num_of_delinquencies",
"income_to_loan_ratio",
"utilization_ratio",
"education_encoded",
"employment_encoded"]]

y=no_outliers["loan_paid_back"]
In [33]:
X_train, X_test, y_train, y_test = train_test_split(
    X, y,
    test_size=0.25,
    random_state=15
)
In [34]:
scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)
In [35]:
lr=LogisticRegression()
lr.fit(X_train_scaled, y_train)
Out[35]:
LogisticRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
LogisticRegression()
In [36]:
y_pred = lr.predict(X_test_scaled)
y_prob = lr.predict_proba(X_test_scaled)[:, 1]
In [37]:
In [81]:
X = sm.add_constant(X)
model = sm.Logit(y, X)
result = model.fit()

print(result.summary())

Optimization terminated successfully.
         Current function value: 0.334290
         Iterations 7
                           Logit Regression Results                           
==============================================================================
Dep. Variable:         loan_paid_back   No. Observations:                19076
Model:                          Logit   Df Residuals:                    19058
Method:                           MLE   Df Model:                           17
Date:                Tue, 13 Jan 2026   Pseudo R-squ.:                  0.3312
Time:                        21:53:51   Log-Likelihood:                -6376.9
converged:                       True   LL-Null:                       -9534.3
Covariance Type:            nonrobust   LLR p-value:                     0.000
========================================================================================
                           coef    std err          z      P>|z|      [0.025      0.975]
----------------------------------------------------------------------------------------
const                   -3.6606      0.417     -8.786      0.000      -4.477      -2.844
monthly_income        7.139e-05   2.61e-05      2.737      0.006    2.03e-05       0.000
debt_to_income_ratio    -7.0442      0.220    -32.006      0.000      -7.476      -6.613
credit_score             0.0112      0.000     26.702      0.000       0.010       0.012
loan_amount           7.049e-06   1.57e-05      0.448      0.654   -2.38e-05    3.79e-05
interest_rate            0.0124      0.012      1.050      0.294      -0.011       0.036
loan_term               -0.0028      0.004     -0.685      0.493      -0.011       0.005
installment             -0.0002      0.001     -0.484      0.629      -0.001       0.001
num_of_open_accounts     0.0042      0.010      0.415      0.678      -0.016       0.024
total_credit_limit   -2.945e-06   2.61e-06     -1.128      0.259   -8.06e-06    2.17e-06
current_balance       5.251e-07   3.71e-06      0.142      0.887   -6.74e-06    7.79e-06
delinquency_history     -0.0423      0.036     -1.163      0.245      -0.114       0.029
public_records           0.0089      0.081      0.109      0.913      -0.150       0.168
num_of_delinquencies     0.0361      0.033      1.091      0.275      -0.029       0.101
income_to_loan_ratio    -0.0005      0.001     -0.357      0.721      -0.003       0.002
utilization_ratio       -0.0053      0.015     -0.348      0.728      -0.036       0.025
education_encoded        0.0362      0.020      1.778      0.075      -0.004       0.076
employment_encoded      -0.9549      0.016    -58.631      0.000      -0.987      -0.923
========================================================================================
In [38]:
print("\nROC-AUC Score:")
print(roc_auc_score(y_test, y_prob))

ROC-AUC Score:
0.8551456344480971
In [ ]: